The research discusses new methods that allow attackers to read from and write to specific parts of high-performance CPUs, like the path history register (PHR) and prediction history tables (PHTs), leading to attacks that reveal a program's control flow history by recovering a secret image using libjpeg routines and enable detailed transient attacks by extracting an AES encryption key. These findings highlight significant security risks for these systems, emphasizing the importance of addressing vulnerabilities in CPU architectures. The research conducted by Hosein Yavarzadeh from the University of California San Diego sheds light on the growing number of threats leveraging Microsoft API, indicating the need for robust security measures and continuous advancements in encryption technologies to protect sensitive data and information
https://thecyberwire.com/podcasts/research-saturday/330/notes